From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1761562AbYAROP3 (ORCPT ); Fri, 18 Jan 2008 09:15:29 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1758182AbYAROPX (ORCPT ); Fri, 18 Jan 2008 09:15:23 -0500 Received: from one.firstfloor.org ([213.235.205.2]:54505 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757005AbYAROPW (ORCPT ); Fri, 18 Jan 2008 09:15:22 -0500 Date: Fri, 18 Jan 2008 15:18:39 +0100 From: Andi Kleen To: "H. Peter Anvin" Cc: Andi Kleen , Kyle McMartin , tglx@linutronix.de, mingo@redhat.com, linux-kernel@vger.kernel.org Subject: Re: [PATCH] x86: make clflush a required feature on x86_64 Message-ID: <20080118141839.GA16738@one.firstfloor.org> References: <20080118045921.GB974@phobos.i.cabal.ca> <20080118060124.GC1726@phobos.i.cabal.ca> <479046BA.7090903@zytor.com> <20080118065428.GA12594@one.firstfloor.org> <4790B01B.8020406@zytor.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4790B01B.8020406@zytor.com> User-Agent: Mutt/1.4.2.1i Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jan 18, 2008 at 08:56:43AM -0500, H. Peter Anvin wrote: > Andi Kleen wrote: > >>Simulators can be fixed, > > > >They could, but why? I don't know of a good reason to require CLFLUSH. > > Well, simulators are generally expected to follow the architecture, not > vice versa. I would tend to agree with the coupling that recent > versions of Bochs appeared to have made here -- I think we're unlikely > to see any processors with sse2 sans clflush, so keeping code branches > in which will never be executed seems like a bad idea in the long term. Here's another argument: Ingo just asked me to add a noclflush option to the code. Guess what check that option will need? Besides compared to the cost of a flushing clflush the branches are absolutely in the noise. -Andi